Nonnative Trout Invasions Combined with Climate Change Threaten Persistence of Isolated Cutthroat Trout Populations in the Southern Rocky Mountains

Effective conservation of Cutthroat Trout Oncorhynchus clarkii lineages native to the Rocky Mountains will require estimating effects of multiple stressors and directing management toward the most important ones. Recent analyses have focused on the direct and indirect effects of a changing climate on contemporary ranges, which are much reduced from historic ranges owing to past habitat loss and nonnative trout invasions. However, nonnative trout continue to invade Cutthroat Trout populations in the southern Rocky Mountains. Despite management to isolate and protect these native populations, nonnatives still surmount barriers or are illegally stocked above them. We used data on the incidence of invasions by nonnative Brook Trout (BT) Salvelinus fontinalis and the rate of their invasion upstream to simulate effects on a set of 309 conservation populations of Colorado River Cutthroat Trout (CRCT) O. c. pleuriticus isolated in headwater stream fragments. A previously developed Bayesian network model was used to compare direct and indirect effects of climate change (CC) alone on population persistence versus the added effects of BT invasions. Although CC alone is predicted to extirpate only one CRCT population by 2080, BT invasions and CC together are predicted to completely extirpate 122 populations (39% of the total) if managers do not intervene. Another 113 populations (37%) will be at risk of extirpation after CC and invasions, primarily owing to stochastic risks in short stream fragments that are similar under CC alone. Overall, invasions and CC will reduce the number of stream fragments that are long enough to buffer CRCT populations against negative genetic consequences and stochastic disturbances by 48, a decrease of 38% compared to CC alone. High priorities are (1) research to estimate how CC and human factors alter the incidence and rate of BT invasions and (2) management to prevent new illegal introductions, repair inadequate barriers, and monitor and address new invasions. *Corresponding author: [email protected] Received March 21, 2016; accepted November 9, 2016 314 North American Journal of Fisheries Management 37:314–325, 2017 © American Fisheries Society 2017 ISSN: 0275-5947 print / 1548-8675 online DOI: 10.1080/02755947.2016.1264507 Conserving native trout requires understanding their ecological relationships at a hierarchy of scales from stream reaches to river basins and managing for regional persistence in the face of multiple stressors. The principal stressors that have caused native trout declines in many regions over the last 150 years are loss of habitat from human land uses and water abstraction, as well as invasions by nonnative trout (Young 1995; Kitano 2004; Hudy et al. 2008). Overlain on these stressors in recent decades has been the influence of climate change, which includes direct effects on water temperatures and stream flows and indirect effects ranging from habitat fragmentation to altered hydrologic regimes (Rahel and Olden 2008; Wenger et al. 2011; Isaak et al. 2012). Effective conservation of native trout will require planning for a future that includes all these factors and taking actions that are both targeted at the most important stressors and effective at ameliorating them (Fausch et al. 2009; Isaak et al. 2015). One group of native trout for which these multiple stressors pose strong threats are the lineages of Cutthroat Trout Oncorhynchus clarkii in the southern Rocky Mountains (Behnke 1992; Metcalf et al. 2012). In this region, the remaining populations are restricted to short headwater stream fragments (often less than 6 km long) by previous habitat loss and invasions by nonnative trout (Harig et al. 2000; Alves et al. 2008; Hirsch et al. 2013). Our recent analysis for 309 conservation populations that remain of the most widespread subspecies, the Colorado River Cutthroat Trout O. c. pleuriticus (CRCT), predicted that the majority of these small populations (63%) are susceptible to extirpation by stochastic disturbance events such as wildfire, sediment debris flows, drying, and freezing, events which are predicted to increase in frequency with climate change (Roberts et al. 2013). In contrast, few of these populations are susceptible to extirpation by future temperature increases alone because nearly all are located in high-elevation headwater streams that are predicted to be either within optimum temperature limits for reproduction and growth or too cold despite the changing climate. Most Cutthroat Trout populations in the southern Rocky Mountains are also protected from upstream invasions by natural or artificial barriers that form their downstream limit (Young et al. 2002; Pritchard and Cowley 2006; Young 2008). Nevertheless, invasions are a regular occurrence, either because barriers are ineffective or fail (Thompson and Rahel 1998; Harig et al. 2000), or humans deliberately move trout above them (Fausch et al. 2009). These nonnative trout can rapidly colonize upstream (Peterson and Fausch 2003), further reducing the length of stream fragments available to support the native trout, and often extirpate Cutthroat Trout within about a decade from many small streams (Peterson et al. 2004, 2008). Broad-scale analyses of climate change effects on native Cutthroat Trout (Williams et al. 2009) and the effects of climate change combined with nonnative trout (Wenger et al. 2011; Isaak et al. 2015) have been conducted for the Rocky Mountain region. However, most analyses to date used air temperature as a surrogate for water temperature when predicting future conditions, and all have used broad-scale species distribution models to predict suitable habitats for native and nonnative trout, making it difficult to identify effects at local scales. In contrast, to be effective for managing the small, restricted Cutthroat Trout populations that remain in the southern Rocky Mountains, analyses will need to account for the unique attributes of the native trout lineages and nonnative trout invasions in the region, and water temperatures and fragment lengths in the specific streams that support conservation populations. Toward that end, our goal was to analyze the combined effect of invasions by nonnative trout and the direct and indirect effects of climate change for the 309 conservation populations of native CRCT throughout the upper Colorado River basin. We ask how important are potential nonnative trout invasions above barriers when combined with the effects of climate change in extirpating native trout and reducing persistence of the remaining populations. Lastly, we consider what management actions are likely to be most effective at ameliorating threats from these multiple stressors to native Cutthroat Trout populations in the southern Rocky Mountains. METHODS CRCT population database.—We used the CRCT Conservation Team database to map the distribution and status of CRCT populations on the landscape (Hirsch et al. 2006; Figure 1). This database combines all available population surveys from management agencies (state, federal, and tribal) and university researchers to specify which stream segments in the National Hydrography Dataset Plus (NHDPlus; 1:100,000; http://www.horizonsystems.com/nhdplus) are occupied by CRCT populations, and their conservation status. We restricted our analysis to CRCTconservation populations, defined as those with ≥90% genetic purity. Most conservation populations were also isolated from nonnative trout invasion by barriers or unsuitable habitat, and free from disease (Hirsch et al. 2006, 2013). We used the same 309 populations analyzed in previous work (Roberts et al. 2013) to allow comparing the effects of climate change alone and combined with nonnative trout invasion. This was the set included in the most current database available in 2009 when we began our previous research. Incidence of nonnative trout invasion.—We restricted our analysis to invasions by Brook Trout Salvelinus fontinalis (BT) because it is the most common nonnative trout directly downstream from the barriers that isolate CRCT populations in the Colorado River basin (Fausch 1989; Young 1995), and extirpation of CRCT populations after BT invasions is welldocumented (Peterson et al. 2004, 2008; Young 2008). Other nonnative trout such as Rainbow Trout O. mykiss and Brown Trout Salmo trutta tend to occur farther downstream, so they invade habitat occupied by CRCT less frequently, except when stocked there. THREATS TO ISOLATED CUTTHROAT TROUT POPULATIONS 315